I collected several datasets that might be useful for these projects, which can be found in Github here:
https://github.com/joeclark-phd/namegen_tests. Full credit: most of them came from another repo by Github user "Tw1ddle". They include such diverse data sets as the names of breakfast cereals, car brands, pokemon, etc. I won't post them all here, but if anyone is interested you may go ahead and prune and prepare some of those datasets (they'll need capitalization and spaces restored, etc.).
One I added myself is a data set I just couldn't pass up. I came across a scholarly article by someone who found a 15th-century manuscript listing "The Names of All Manner of Hounds", an inventory of 1064 names for dogs from the late medieval era. You can read the article here:
https://www.brepolsonline.net/doi/10.1484/J.VIATOR.1.103488I copied over all the dog names into the file attached. Here's just a sample:
Argente
Aldirman
Archere
Archebawde
...
Beawyew
Blamer
Bragger
Braynesike
...
Cannone
Cradokke
Charlemayne
Creseyte
...
SJW: Added the hound names for v2.6